Fall 2022

Quantifying Uncertainty: Stochastic, Adversarial, and Beyond


Laura Doval (Columbia University), Nicole Immorlica (MSR NYC), Thodoris Lykouris (MSR NYC)

The workshop will explore online decision-making under different modeling assumptions on the reward structure. The two classical approaches for that consist of the setting where rewards are stochastic from a distribution and the one where they are adversarially selected. We will discuss different hybrid models to go between these extremes (data-dependent algorithms that adapt to “easy data”, model-predictive methods, ML-augmented algorithms, etc). We will also consider settings where the rewards come from agents with particular behavioral or choice models and how the algorithms need to change to adapt to that.

Further details about this workshop will be posted in due course. Enquiries may be sent to the organizers at this address.